DOI: 10.11647/OBP.0029.03
Right now I am writing a philosophical essay about the sociology essay “On Doing ‘Being Ordinary’”, by Harvey Sacks.1 The thesis of this brilliant essay (Sacks’s, not mine) is that no matter what we do, we are doing something else in addition, namely, being ordinary. By “being ordinary”, Sacks means doing something that is ordinarily done in a situation like ours, conceived as a situation ordinarily encountered by people like us, conceived as people of some ordinary kind. I am a philosophy professor, an ordinary sort of person to be. Even if I held the Extraordinary Chair in Philosophy (I don’t), such a chair would be an ordinary sort of position for a professor to hold. And an ordinary sort of thing for a professor to do is to write an essay about a topic that he finds somewhat out of the ordinary, though not too far out.
Sacks puts his point like this:
[T]here is an infinite collection of possibilities, of things to do, that you could not bring yourself to do. In the midst of the most utterly boring afternoon or evening you would rather live through the boredom in the usual way — whatever that is — than see whether it would be less or more boring to examine the wall or to look in some detail at the tree outside the window. (415–416)
In other words, there are ordinary ways of doing “being bored” — flipping unseeingly through an old magazine, staring unhungrily into the fridge — and when you are bored, you do it in one of those ways.
Another example:
Among the ways you go about doing “being an ordinary person” is to spend your time in usual ways, having usual thoughts, usual interests, so that all you have to do to be an ordinary person in the evening is turn on the TV set.
Now, the trick is to see that it is not that it happens that you are doing what lots of ordinary people are doing, but that you know that the way to do “having a usual evening,” for anybody, is to do that. (415)
The reference to TV in this passage may seem to suggest that Sacks is talking about being mundane or even inane. But he thinks that the task of being ordinary confronts you in the course of even the most exotic episodes:
[W]hether you were to have illegitimate experiences or not, the characteristic of being an ordinary person is that, having the illegitimate experiences that you should not have, they come off in just the usual way that they come off for anybody doing such an illegitimate experience.
When you have an affair, take drugs, commit a crime, and so on, you find that it has been the usual experience that others who have done it have had. […] Reports of the most seemingly outrageous experience, for which you would figure one would be at a loss for words, or would have available extraordinary details of what happened, turn out to present them in a fashion that has them come off as utterly unexceptional. (418)
Although Sacks speaks here of “having experiences”, the passage makes clear that he means “exploits” or “adventures”. Sacks is saying that if you pass up an evening of TV in order to rob an all-night grocery, you will still do “robbing an all-night grocery” in the ordinary way. (You’ve seen it on TV.)
A Version of Moral Relativism
In order for anyone to aim at doing what’s ordinary, there has to be something that is ordinarily done, which is whatever is done by others, who, according to Sacks’s thesis, are also aiming to do what’s ordinary. Hence everyone has to converge on a repertoire of ordinary actions that isn’t defined in advance of everyone’s converging on it. Ordinariness is socially constructed, and constructing it is a classic coordination problem.
Because ordinariness is socially constructed, it is also local, in the sense that it is relative to some population of agents who interact regularly, usually because they live in one another’s vicinity. What’s ordinary in New York or Omaha is not the same as what’s ordinary in Ramallah or Singapore, as everyone knows. One way to think of this phenomenon would be to imagine a domain of actions — or, more accurately, action-types — from which different communities2 select different subsets on which to converge as the ordinary set. I will argue that this conception of the coordination problem would be mistaken, because there is no neutral domain of actions from which a community can select. In constructing what’s ordinary, a community also constructs the domain of action-types.
The result is that communities can find themselves unable to disagree about what should be ordinarily done, because they differ with respect to what is doable: there is no neutral domain of action-types from which they choose what to do. What’s more, action-types are invented, and there is no domain of inventable action-types from which communities can choose which ones to invent, much less disagree about such choices. Insofar as they can disagree about which action-types to invent, they disagree just by living differently, each converging on ordinary choices from among its own, socially constructed domain of doables.
This obstacle to disagreement extends to morality. Disagreement about morality is disagreement about what may or may not be done, and so it requires agreement about what is doable. For communities with different domains of doables, the question what may or may not be done is therefore moot.
Now, the nature of moral disagreement is the issue between moral relativists and their critics. Some think that the issue, more specifically, is whether moral disagreement can be faultless, in the sense that both parties to the disagreement are right. In my view, moral relativists should not rest their case on the possibility of faultless disagreement; they should rest it instead on the impossibility of disagreement altogether. Both of two parties can be entitled to stick to their moral views not because both are right but rather because it is a moot question which one is right, so that there is no moral question to adjudicate. Relativism requires only that there be no judicable question between moral views.
My goal in this chapter is to explain the construction of different practical domains by different communities. My explanation will be that the social construction of doables is governed by the same forces that produce the phenomenon of “doing ‘being ordinary’”, as described by Sacks. I’ll conclude by returning to the implications for moral relativism.
Sacks belonged to a group of sociologists who adopted the name of ethnomethodologists, because they sought to catalog the “methods” of ordinary life. As the name suggests, ethnomethodologists modeled themselves on anthropologists, for reasons explained by one of Sacks’s co-authors, Emanuel Schegloff:3
[A]nthropologists had had good reason over the course of the development of their discipline to wonder what some “natives” were “doing” by conducting themselves in a certain way, for example, by talking in a certain way. Both linguistic obstacles and so-called “culture differences” could pose quite sharply the problem of “recognizing actions,” and then the analytic problems of describing what those actions were and how they were done or accomplished. […]
[…] [T]he same sort of attention to, and description of, actions in one’s own culture never had quite the same resonance, even in anthropology. The actions were “transparent” to comembers of the culture, even naming them by action names might appear a bit arch and scholastic. Even more so was this treatment accorded accounts of the practices, rules, or mechanisms by which these actions were done. They just “were” invitations, requests, promises, insults, and so on.
This passage should ring a bell for philosophers, who know that we never detect what “just is”. Our observations are informed by concepts of what there is to be observed — concepts of observables. Schegloff was pointing out that we act under concepts of what there is to do — concepts of doables.4
There is an important difference between observables and doables, however. The fact that observation is informed by concepts does not rule out the possibility that some of those concepts are more faithful than others to the structure of the world: there may be better and worse ways of articulating what we see, because the world itself may be articulated — may have joints at which some concepts carve it better than others. By contrast, which actions there are depends on which action concepts figure in people’s intentions. We talk about “taking” an action, as if we were picking an apple from a tree, but actions don’t antecedently exist in nature, waiting to be picked. What we call taking an action is actually making an action, by enacting some act-description or action concept. Which actions we can make depends on which descriptions or concepts are available for us to enact.
Individuals can sometimes invent new things to do, but invention is the exception rather than the rule. An agent cannot invent an entire ontology of actions from scratch; for the most part, he must choose from a socially provided repertoire of action concepts. Just as he sees things of kinds that he has been taught can be seen, so he does things of kinds that he has been taught can be done. As the sociologist Ann Swidler has argued, a person’s culture provides him not so much with values and ends as with a “toolkit” of possible actions:5
People do not build lines of action from scratch, choosing actions one at a time as efficient means to given ends. Instead, they construct chains of action beginning with at least some pre-fabricated links. Culture influences action through the shape and organization of those links, not by determining the ends to which they are put.
These socially provided links for possible chains of action are what I interpret Sacks to mean when he speaks of what is ordinary, and they are what I will mean by ‘doables’.
It is not just a shortage of time or energy or imagination that prevents an individual agent from venturing outside the predefined range of doables. The shared ontology facilitates mutual understanding and cooperation. This point was emphasized by Alfred Schutz, a philosopher from whom the ethnomethodologists drew inspiration.6
On the one hand, I have — in order to understand another — to apply the system of typifications accepted by the group to which both of us belong [ …]. On the other hand, in order to make myself understandable to another, I have to avail myself of the same system of typifications as a scheme of orientation for my projected action.
What Schutz calls a “typification” is simply the concept of a typical action — something ordinarily done. One draws on a shared system of typifications, according to Schutz, in order to understand what others do and to do things that others can understand.7
That taxonomies of doables are socially constructed can be inferred from cultural differences in action concepts. A commonplace among linguists, for example, is that some cultures have only one verb for the actions that we discriminate as eating and drinking;8 indeed, some have a single verb for eating, drinking, smoking, and kissing.9 In such cultures, taking in through the lips is a more salient action-type than taking in smoke; in our culture, taking in smoke is more salient: we don’t even have a verb for the other.
Social construction extends even to the level of bodily movements, for the simple reason that parts of the body are differently individuated in different societies.10 The social anthropologist Edwin Ardener offers an example of how different taxonomies of the body generate different taxonomies of action:11
Let us consider the shaking of hands in England and among the Ibo of south-eastern Nigeria. In both languages there are apparently intertranslatable terms for the gesture (Ibo ji aka). Although aka is usually translated “hand” the boundaries of the parts concerned are, however, quite different. The English “hand” is bounded at the wrist. The Ibo aka is bounded just below the shoulder. […] The more open-gestured nature of the Ibo handshake compared with the English handshake is linked in part to this difference of classification. For the English-speaker the extreme, “formal” possibility of presenting an only slightly mobile hand at the end of a relatively stiff arm becomes a choice reinforced by language. For the Ibo-speaker, even if that is a possible gesture it has no backing from language.
Consistently shaking hands alone, with articulation only at the wrist, might therefore seem to the traditional Ibo a slightly incomprehensible restriction of movement, equivalent perhaps in flavour to being, in the English case, offered only two or three fingers to shake. From the opposite point of view, to the English speaker “shaking hands” and “arm-grip” are two kinds of greeting. To the Ibo they are degrees of intensity, demonstrativeness, of warmth, of “the same” greeting.
Ardener goes on to explain that although helping can be described in both languages as “lending a hand”, requests for a helping hand are met differently. In circumstances that call for the literal proffer of a hand, an Ibo may proffer a forearm even if his (English) hand is available. The way an Ibo lends a hand may therefore strike an English speaker as cold — he wonders, “Why won’t you give me your hand?” — which is how the English speaker’s handshake strikes an Ibo, who wonders, “Why do you give me just the tip of your hand?”
Ardener suggests that even the categories of action and behavior are socially constructed. Among the Ibo, he claims, all generic terms for behavior carry connotations of social approval or disapproval. So there is only good or bad behavior for the Ibo, not behavior simpliciter.12
The Social Construction of Speech Acts
Consider next the example of speech acts. I choose speech acts for two reasons. First, they are actions that have been extensively taxonomized by philosophers, in speech-act theory. Second, the hierarchy of speech acts can seem to be universal, at least in its overall shape, and so the work of constructing it is easily overlooked. Although the acts of filibustering and excommunicating are obviously constituted by social practices, the acts of asking and asserting do not immediately strike us as culture-bound. Indeed, John Searle claims that the fundamental illocutionary forces — the assertive, the directive, the commissive, the declarative, and the expressive — are natural kinds that exhaust the possible uses of language.13 In other words, the fundamental doables of verbal communication are simply there to be done, according to Searle. Yet it turns out that Searle’s inventory of illocutionary forces may be culture-bound after all, as is our taxonomy of the acts that exert those forces.
Commissives among the Ilongot
Searle’s inventory of illocutionary forces was first challenged by the anthropologist Michelle Rosaldo in an account of speech acts among the Ilongot people of the Philippines.14 According to Rosaldo, neither the expressive nor the commissive categories have clear instances among the Ilongot. The closest approximation to commissives such as promises among the Ilongot neither express the same attitudes nor incur the same consequences as promises in English. They are rather oaths nominating calamities to befall the speaker if he fails to follow through. Such failures do not occasion guilt or apology:
Repeatedly, I was outraged to find that friends who had arranged to meet and work with me did not appear at the decided time — especially as they would then speak not of commitments broken, or of excuses and regrets, but of devices (such as gifts) that might assuage the generally unexpected and disturbing anger in my heart. To them, it mattered that I was annoyed (a dangerous and explosive state), but not that someone else, in carelessness, had hurt and angered me by failing to fulfill commitments I had understood as tantamount to promises. (218)
In this culture, the way to give people grounds for counting on one’s future action is to make oneself vulnerable to harms if one should disappoint them. One alters one’s own incentives, in other words, without making any commitment to others, so that if one lets them down, one will need only to watch out for oneself and to assuage their disappointment.
Evidentials
Whereas the Ilongot fail to exercise illocutionary forces that are included in Searle’s inventory, other communities exercise illocutionary forces that Searle’s inventory lacks. Consider languages that use grammaticized rather than lexical evidentials — that is, grammatical means of indicating the speaker’s epistemic relation to what is said. In these languages, the speaker’s epistemic position is indicated by word-endings rather than by words or phrases. One ending may be used for statements of what the speaker has witnessed firsthand, another for what the speaker has heard from others, a third for what is generally known in the community, a fourth for what the speaker infers or intuits. In some languages, such evidentials are mandatory; that is, a speaker must always add a word-ending that indicates his epistemic position.
Whereas an English speaker says, “He’s out hunting, I’m told”, a speaker of Quechua says, “He’s out hunting-si”, ‘-si’ being the reportative evidential in his language.15 In the Amazonian language Matses, evidentials are mandatory in all past-tense statements, which include all statements about absent parties.16 If a man is asked how many wives he has and they aren’t present, he will answer with a statement whose grammatical evidentials can be translated into English only with lexical evidentials, like this: “There were two (last time I checked).”17
The first thing to say about languages with mandatory evidentials is that they do not have a speech act corresponding to our bare assertion. It is impossible, in these languages, to put forward a proposition as true without indicating how it is known to be true. In place of our bare assertion, they have several speech acts, which we might call testifying, reporting, speculating, and the like.
By the same token, we do not have speech acts afforded to speakers by languages with evidentials. We can of course say that what we are asserting is known firsthand, by intuition, or whatever, but we thereby insert material about our sources into the truth-conditions of our statement, whereas evidentials need not contribute to truth-conditions. When an English speaker adds the words “I’m told” or “reportedly” to an assertion, he not only indicates his epistemic relation to what is asserted; he also implies that a report of it has been made. One can therefore challenge him by saying, “That hasn’t been reported.” By contrast, a reportative evidential in Quechua can mark a statement as hearsay without yielding the further implication that it was heard, and so the challenge “That hasn’t been reported” would misfire. The reportative evidential thus makes possible a speech act that is not available to speakers of English.
Indeed, it has been claimed that the reportative evidential in Quechua produces a speech act whose illocutionary force does not fit into Searle’s taxonomy. In addition to leaving the truth-conditions of a statement unchanged, an evidential can lack the force of an epistemic modal. In Quechua, “He’s out hunting-si” indicates a lack of certainty on the part of the speaker, as would the English statement “He may be hunting”, but the Quechua statement is co-assertible with “He isn’t hunting”, whereas “He may be hunting” is not. And unlike “He may be hunting”, the Quechua statement is co-assertible with “I know he isn’t hunting” as well. Thus, Quechua speakers can say “He’s out hunting” without implying that it may be true, that it is compatible with what is known, or that it has been heard, and yet in a tentative register reserved for hearsay. An evidential like the Quechua ‘-si’ can thus generate a speech act that has the illocutionary force of “presenting” a proposition without vouching for its truth, actual or epistemically possible — an illocutionary force that isn’t accommodated by Searle.18 Once again, the individuation of speech acts turns out to be culture-bound.19
Grice’s maxims
Genres of assertion unknown to English speakers also appear in cultures where Grice’s conversational maxims are said not to hold. In Malagasy society, for example, assertions are designed to be strategically uninformative, for various cultural reasons.20 Life is generally lived in the open, and exclusive knowledge is therefore a rare commodity, possession of which confers prestige on the knower. There is also a risk of losing face if one’s claims turn out to be false. Consequently, speakers go out of their way to convey less information than they could conveniently convey. Asked where someone is, a speaker will say, “In the market or at home”, despite knowing which is the case.21 Asked when someone will be at home, a speaker will say, “If you don’t come after five o’clock, you won’t find him”, a circumlocution designed to be less informative than saying that the person will be in after five. Speakers use indefinite descriptions for people whose identity they know, so that a speaker who sees his own mother at the door may say, “Someone is here.”22 Hence “Someone is here” does not implicate that the speaker doesn’t know who.
A more extreme phenomenon has been observed in cultures where the paramount values are honor and prestige.23 According to an ethnography of an Egyptian village, conversation among the inhabitants is often a matter of negotiating a common ground that is more attuned, in the first instance, to social norms other than truthfulness:24
Fairly transparently false assertions can still be effective in shaping the future course of a conversation because in most cases they are immune from challenge for reasons of politeness — showing respect not only for the other’s honesty but also for their authoritativeness and soundness of judgement. There may be mutual knowledge that an assertion is false, while yet speakers cooperate on treating it as if true, or at least valid, for immediate conversational purposes (‘valid’ as with opinions one respects but disagrees with). But there has to be left open some possibility, however remote, that the assertion could be true, for the pretence to be workable; otherwise it slips over into irony [… ].
These regularities reveal that villagers are — and need to be — highly sensitive to issues of truthfulness and the concomitant sanctions, and to where truthfulness is and is not to be expected. They have a name for the type of speech where it is not expected: kala:m ‘(mere) words/talk’, and it is freely acknowledged as occurring in a wide range of situations.
Kala:m is the name of an assertoric genre that does not seriously aim to be true but is not fictional, either, and is nevertheless permitted.25 Conversely, the Trobriand Islanders have a formula that means “Now I’m speaking the truth” — a formula that is not usually invoked, so that most indicative utterances can be retracted, if necessary, as having been sopa, or unserious.26
Untruthfulness
Among the Mopan Maya, the only word translatable as “lying” is tus, which carries no connotation as to the speaker’s knowledge or intention:27
There exists no other candidate lexeme in Mopan for the notion of “lying” or “stating falsehood,” and the translation “lies, lying” is the only one ever offered for this form by bilingual Mopan speakers. Harshly or mildly applied, a negative connotation is always present to some degree in uses of this word. A characterization of another’s utterance as tus, however, is based exclusively on the perceived truth value of expressions and not on the intentional or belief states of the speaker. This is so even when the speaker merely translates the opinions or repeats the words of another […]. Accordingly many cases of expressions that might be categorized elsewhere as “errors” are condemned in Mopan as tus.
According to Clifford Geertz, the Javanese use the word étok-étok to mean “proper lying”, which is not quite the same as our “white lie”. An informant explained it to him like this:28
He said: “Suppose I go off south and you see me go. Later my son asks you: ‘Do you know where my father went?’ And you say no, [you] étok-étok [that] you don’t know.” I asked him why should I étok-étok, as there seemed to be no reason for lying, and he said, “Oh, you just étok-étok. You don’t have to have a reason.”
Geertz elaborates:
When we tell white lies, we have to justify them to ourselves even though the justification be weak. […] For the Javanese […] it seems, in part anyway, to work the other way around: the burden of proof seems to be in the direction of telling the truth. The natural answer to casual questions, particularly from people you do not know very well, tends to be either a vague one (“Where are you going?” — “West”) or a mildly false one; and one tells the truth in small matters only when there is some reason to do so.
In some Lebanese circles, speakers of Arabic puff themselves up, and put others down, by passing off out-and-out falsehoods, which are admitted after the fact to be lies, or kizb.29 In order to have the desired effect on the relative status of speaker and hearer, these falsehoods must eventually be revealed, and so they are not exactly the same as lies in the English-speaker’s sense, but because they are so consequential in social terms, they are not exactly the same as “leg-pulling” or “April Fools’” jokes, either.30
Russian has two words for lying: lozh, which denotes an out-and-out lie, and vranyo, for which English has no equivalent. Vranyo is not exactly bullshitting, not exactly fibbing, not exactly joshing, not exactly telling tales. Here is an example:31
I was once present when a USSR-domiciled Russian visitor to England spoke to a British host about a Russian émigré known to both of them. According to the visitor, one of the émigré’s sons had recently returned to Moscow after a stay in Paris; had published a book of which the exact title was given; had changed his name; had undergone various other adventures. The host listened with a straight face, thus preserving the conventions, though he was a close friend of the family concerned and knew that every word was untrue. Did the narrator realise that he had been identified as purveying vranyo? Yes and no.
The purveyor of vranyo does not quite expect to be believed. He does count on not being unmasked — on receiving a straight-faced hearing — but he also prefers aesthetic appreciation to naive credence.32 It is even unclear whether he thinks that he is telling untruths. Dostoyevsky put it like this: “You have told such fantastic stories […] that, though you have started to believe in yourself half-way through your story (for one always does begin to believe in oneself half-way through a story), nevertheless when you go to bed at night and have enjoyable memories about the pleasant impression made on your listener, you suddenly pause and remark involuntarily, ‘Heaven, what rubbish I talked.’”33
Now, you may say, “But kala:m, kizb, étok-étok, and vranyo are just practices that we English-speakers don’t happen to have. So what?” Here’s what: These practices affect the individuation of speech acts. For both speakers and hearers of Russian, there are two ordinary types of untruthfulness, both of which would count as lying in English. If the Russians’ vranyo is a type of action that we don’t happen to have, then by the same token our lying is an action-type that they don’t quite duplicate, either.34 The same point applies to the Javanese, with the additional note that étok-étok and vranyo are different from each other, too.
I would argue that these forms of untruthfulness also strain Searle’s taxonomy of illocutionary forces. In the practices that I have just described, propositions expressed in indicative sentences are not put forward simply as true or as to be believed. In some cases, they are put forward as entries into a common ground that may have no use beyond the present interaction. In other cases, they are moves in a conventionally sanctioned bluffing game, put forward as maybe true, maybe false, to be believed at one’s own risk. I question whether these utterances have assertoric force.
There may even be cultural differences in the conceptualization of null speech acts — that is, acts of not speaking. In Anglo-American culture, silence is strenuously avoided in social situations, especially among strangers newly introduced, but for the Western Apache, “silent co-presence” is a way of socializing, especially among strangers, so that hosts and guests may sit silently for the first half-hour of a visit.35 This community affords its members an act of sociable silence that is not available to Anglo-Americans.
One wonders, then, whether the Western Apache lack a precise correlative to the Anglo-American action-type of speaking, just as the Ibo lack a correlative to mere behavior. Or perhaps it is only speaking up that has no correlative among the Apache — at least, none without a tinge of the Anglo-American interrupting.
Why Construct Ordinariness?
Of course, part of the reason why speech is constrained by the local repertoire of speech acts is that speech acts call for uptake: you cannot use Quechua evidentials with hearers who don’t understand the language. But the fact, which was noted by Schutz, is that you need uptake for most of what you do. It’s imperative that others be able to interpret not just your speech but your behavior in general, so that they know how to interact with you — or how to avoid interacting with you, for that matter.
Navigating a crowded sidewalk requires you to let others understand what you are doing and even what you are feeling. Are you strolling aimlessly, rushing for a light, pausing at a store window, stopping to beg? All of these actions belong to a socially constructed ontology that must be shared by agent and interpreter if mutual understanding is to be attained.
If you want to ride the subway in New York, you’d better know how to make yourself understood — not necessarily in English, which is of limited utility, but in the behavioral vocabulary of crowding in, brushing past, reaching around, stepping aside; sprinting as opposed to chasing; nudging as opposed to jostling. When you come to New York, speak the local language of action: do not improvise.
Think of the situation in which you realize that you have been staring at a point in space that has turned out to be occupied by someone else’s face. You have carelessly let him think that you were staring at him, and now you have to convey that you weren’t staring, because you didn’t even see him, though you obviously saw him think that you were staring, and so you somehow have to convey that his thinking so was the first thing you saw. Then again, whether the person thinks you are staring at him will depend on whether he thinks that he has somehow led you to think that he is seeking attention, since an invited gaze is not a stare. If he thinks he might have seemed to be trying to catch your eye, then he may wonder whether you are paying attention rather than staring — in which case, he will make a show of staring at his feet, so as to convey that he isn’t putting on a show, except for that one. The entire interaction requires a shared taxonomy of gazes.
Joint intentions
The need for a shared taxonomy of actions is intensified by the prevalence of joint intentions in human affairs.36 As Margaret Gilbert has explained, you and I are in a position to speak of “our” doing something only if each has formed an intention that’s conditional on the other’s and only if we have common knowledge of those intentions.37 Each must intend to do his part provided that the other intends likewise, and both must know that they have these intentions, and know that they know, and so on. This configuration of attitudes is most easily brought about by an exchange of words — “I’m willing if you are”, “Then I’m willing, too” — but it can also be brought about tacitly, and tacit joint intentions are virtually ubiquitous, even where the resulting collaboration is not evident.
When I walk on a city street, I intend to leave the other pedestrians alone, provided that their intentions toward me are similar. I do not intend to leave them alone if they intend to interfere with me, nor even to leave them alone if they intend to leave me alone no matter what my intention toward them might be. If they intended to leave me alone even if I decided to stare at them, or to take shelter under their umbrellas, or to dart into a taxi they have hailed, I might be tempted to do those very things. My intention toward them is therefore contingent on their having a similarly contingent intention toward me.
These intentions are common knowledge among us. We haven’t exchanged explicit commitments: we haven’t said, “I’m willing to leave you alone, provided that you intend likewise toward me.” But we have implicitly signaled our intentions by our behavior. We avoid looking at one another too long, standing too near, following too closely, speaking or gesturing in one another’s direction, and we do so in a way that is both defensive and deferential. None of us is certain of the others’ intentions, but we are fairly confident, and our confidence is justified, with the result that, if vindicated in the event, it will turn out to have constituted common knowledge. We therefore satisfy the conditions for jointly intending to leave one another alone. Whereas some people jointly intend to walk together, we pedestrians jointly intend to walk apart.
Joint intentions require a shared taxonomy of actions. Indeed, they require a taxonomy that is not only shared but known to be shared — a taxonomy that is common knowledge. Each participant in the intention must know what the other intends; know that the other knows what he, the first knower, intends; and so on. Such knowledge is possible only if there is common knowledge as to the doables that can be intended.
Scenarios
Many of these joint intentions are intentions to participate in socially shared scenarios of standard interactions. Roger Schank and Robert Abelson argue that a robot would need to know many such scenarios in order to simulate an intelligent agent.38
Schank and Abelson’s favorite example is the restaurant scenario, which can be pictured as a flow diagram of how a visit to a restaurant typically unfolds. Either you must wait to be seated — in which case, you may or may not be asked whether you have a reservation — or you are permitted to seat yourself; then you wait until someone brings a menu, or tells you what’s available, or both; then you are left to deliberate; then the server arrives, and the diners take turns stating their choices from the available items, leaving out condiments, which are already on the table, and dessert, which is ordered later; and so on. Even if you knew that restaurants are places to eat, you would have trouble extracting a meal from one of them if you didn’t know how the scenario goes. If you didn’t know the scenario, of course, you might ask for directions at the door, but you would then have to know the “asking for directions” scenario.
To say that you know the scenario is to say that you are following the same flow diagram as everyone else. If you have mistakenly walked into the home of someone holding a private dinner party, you won’t get very far with the restaurant scenario. As soon as the host holds out his hand for a welcoming handshake, you’ll know that something is wrong. When you enter a restaurant, you begin the restaurant scenario because you believe that everyone else will follow the same scenario, and they follow suit because they believe likewise of you.
There are even scenarios for deprecated interactions (Sacks’s “illegitimate experiences”). In order to pull off a mugging, you have to let the victim know he is being mugged, so that he will play his part.39 When he surrenders his wallet and you take it, the two of you will thereby enact a handoff, according to a scenario that you jointly intend to enact: it takes two to hand over a wallet. If you’re the jumpy type, you may shout, “No false moves!”, meaning “Stick to the scenario!” or, as Sacks might put it, “Be ordinary!”
Scenarios for social interaction usually have their own entries in the taxonomy of actions: ordering a meal, welcoming a guest, shaking hands. In order to make clear that I’m speaking of more than the individuation of one-off actions, I’ll adopt the term ‘practical repertoire’ from here on.
Framing
Our practical repertoire shapes our behavior; for as is well known in social psychology, the act-descriptions in which alternatives are framed strongly influence our choices.40 Insofar as we choose among items in the socially shared repertoire, we are under the influence of a socially defined decision frame.
The framing effect is not just psychological but logical. In formal decision theory, acts are represented as choices of outcomes or of gambles on possible outcomes. The things that can be done — the doables — are determined by the outcomes that can be chosen or gambled on. Decision theorists have noted that what counts as rational or irrational, in the terms of their theory, depends on how doables are individuated.
John Broome illustrates this point with the example of an agent who seems to have irrational preferences because he prefers sightseeing over mountaineering, mountaineering over staying at home, and staying at home over sightseeing.41 Broome explains that this agent’s options can be subdivided according to the alternatives rejected, yielding options such as mountaineering-instead-of-sightseeing, mountaineering-instead-of-staying-home, and so on. The agent’s preferences can then be represented as rational, since the act of mountaineering-instead-of-sightseeing may be dispreferred as unsophisticated, the act of staying-home-instead-of-mountaineering dispreferred as cowardly, and the act of sightseeing-instead-of-staying-home dispreferred as over-tiring, given that sophistication and courage are no longer at issue. Broome says that re-description may be appropriate in this case, because it draws distinctions that make a rational difference. As Broome points out, however, we can erase intransitivity in any set of preferences whatsoever by subdividing the options in this fashion, and decision theory cannot distinguish between the cases in which doing so is illuminating and those in which it amounts to sheer gerrymandering. Broome therefore argues that decision theory must be supplemented by “principles of rational indifference”, which will rule out irrelevant distinctions among alternatives.
Yet principles of rational indifference will not help us decide whether to have the concepts of mountaineering and sightseeing to begin with. If people lack the concepts of mountaineering or sightseeing, they lack options for how to spend their spare time, and not for lack of mountains or sights. In order to do what we call sightseeing, they would first have to invent it.42 And then there may be a rational difference, in Broome’s sense, between mountaineering-instead-of-sightseeing and mountaineering in a context where sightseeing is not an alternative. Similarly, there may be a rational difference between truth-telling where vranyo is an alternative and truth-telling where it is not. Inventing an action-type therefore alters the decision frame, and even if there are principles of rational indifference for the resulting frame, there can be no rational principles for whether to invent a new action-type, since inventions are not chosen from among alternatives.
Foundations for Moral Relativism
Do Anglo-American readers of this book have a moral disagreement with people who practice kala:m, étok-étok, kizb, or vranyo? It would seem odd for us to condemn those practices as dishonest. The strongest negative attitude we are likely to have is to be glad that we don’t live among the practitioners, while granting that if we did, we probably wouldn’t regret it. Alternatively, we might feel somewhat envious of the Russians, whose social life is spiced with creative bluffing. In any case, none of these attitudes would support moral agreement or disagreement, despite our own moral seriousness in matters of truth-telling and candor.
The fact is that each community has inherited from its ancestors a vast decision frame consisting in a distinctive taxonomy of actions. An inhabitant of one community can of course consider the option of relocating to another, but he can consider that option only under an act-description available within his own decision frame. There is no neutral act-description under which to choose a community: our emigrating may be no more universal an action-type than mountaineering or sightseeing.
The appearance of inter-community disagreement is often due to comparisons between action-types that are not in fact alternatives. If we Anglo-Americans deprecate étok-étok by saying, “They ought to tell the truth,” we are using a concept of truth-telling that may not be rationally salient, or even available, in a community where étok-étok is a common form of indicative utterance. Étok-étok must be evaluated in the context of the speech acts available within the same decision frame.
In order to evaluate the practice of étok-étok, then, we would have to ask the Javanese why, in a particular situation, they choose étok-étok over some other speech act available to them, a question to which, according to Geertz, the answer would be, “You don’t have to have a reason; you just étok-étok” — at which point we would realize that it’s going to be a long night. The first answer or demurral will lead to another and another, until we have mapped out a very large web of practices and reasons, to which the only possible reaction will be relief or regret that we aren’t Javanese.
The obstacle to disagreement is not that we cannot commensurate between conceptual schemes. Let us grant that the anthropologists have enabled us to understand the range of action-types available to the practitioners of étok-étok. The problem is that if we deprecate étok-étok as against the actual alternatives, we will be engaged in an intra-community disagreement — disagreeing as one Javanese with another — thereby conceding the point that disagreement must take place within a cultural context. And given the framing effect, there is a good chance that we will end up choosing to étok-étok after all.
Footnotes
1Structures of Social Action: Studies in Conversation Analysis, ed. J. Maxwell Atkinson and John Heritage (Cambridge: Cambridge University Press, 1984), Chapter 16.
2While using the term ‘community’, I acknowledge that it is problematic. All of the alternative terms (‘culture’, ‘society’, and so forth) are problematic as well. There aren’t well-defined communities or societies or cultures. When we theorize about them, we are simplifying. But I think that the simplification is warranted as an idealization for the purpose of theorizing.
3Emanuel A. Schegloff, “Confirming Allusions: Toward an Empirical Account of Action”, American Journal of Sociology 102, no. 1 (1996): 162–163.
4On the individuation of options, see also Matthew Noah Smith, “Practical Imagination and its Limits”, Philosophers’ Imprint 10, no. 3 (2010): http://www.philosophersimprint.org/010003/. Of course, it won’t come as news to philosophers that an action is always performed under a description. But the phrase ‘performed under a description’ is potentially misleading, because it means different things to the agent than it does to an observer. An observer sees a bodily movement and considers various descriptions under which it might have been performed. But the agent didn’t choose among different descriptions for that particular bodily movement; he chose among different descriptions to enact. And the bodily movement by which he enacted his chosen description was determined, at the most basic level, by sub-agential skills stored in his brain and body. So even if there is a neutral substrate of movement that can be variously interpreted, that substrate is irrelevant, indeed invisible, from the agent’s point of view; from the agent’s point of view, there are merely act-descriptions to realize.
5“Culture in Action: Symbols and Strategies”, American Sociological Review 51, no. 2 (1986): 277.
6Alfred Schutz, “Equality and the Meaning Structure of the Social World”, Collected Papers II, ed. Arvid Brodersen (The Hague: Martinus Nijhof, 1964), 237. See also “Common-Sense and Scientific Interpretation of Human Action”, Philosophy and Phenomenological Research 14, no. 1 (1953): 19–20, 25–26, and “Concept and Theory Formation in the Social Sciences”, The Journal of Philosophy 51, no. 9 (1954): 268.
7Note that Schutz’s thesis doesn’t apply to solitary actions: you don’t choose recognized ways of being bored in order to make yourself interpretable to others. (If others were around to interpret you, you might not be so bored.) I would say that you choose recognized ways of being bored in order to be interpretable to yourself.
8Anna Wierzbicka, “All People Eat and Drink. Does This Mean That ‘Eat’ and ‘Drink’ are Universal Human Concepts?”, in The Linguistics of Eating and Drinking, ed. John Newman (Amsterdam: John Benjamins, 2009), 65–89. For a general discussion of whether there are lexical universals, see Kai von Fintel and Lisa Matthewson, “Universals in Semantics”, The Linguistic Review 25, no. 1–2 (2008): 139–201.
9William A. Foley, Anthropological Linguistics: An Introduction (Malden, MA: Blackwell, 1997), 28; Borut Telban, “The Poetics of the Crocodile: Changing Cultural Perspectives in Ambonwari”, Oceania 78, no. 2 (2008): 227. As Randall Dipert has pointed out to me, there are certainly universal categories of action. Even if ingesting is not among them, a category such as self-moving, or locomotion, probably is. But such categories are universal in the sense that every community has action concepts that fall within the category, not in the sense that every community has the category itself as an action concept.
10See the special issue on terms for body parts in Language Sciences 28 (2006). See also Drid Williams, “Taxonomies of the Body”, Journal for the Anthropological Study of Human Movement 1, no. 1 (1980): 1–11.
11“Social Anthropology, Language and Reality”, in Semantic Anthropology, ed. David Parkin (New York: Academic Press, 1982), 4–5.
12“‘Behaviour’: A Social Anthropological Criticism”, Journal of the Anthropological Society of Oxford 4, no. 3 (1973): 152–154, p. 153, reprinted in Journal for the Anthropological Study of Human Movement 10 (1999): 139–141, and in Ardener, The Voice of Prophecy and Other Essays, ed. Malcolm Chapman (Oxford: Blackwell, 1989), 105–108.
13John R. Searle and Daniel Vanderveken, Foundations of Illocutionary Logic (Cambridge: Cambridge University Press, 1985): “Illocutionary forces are, so to speak, natural kinds of uses of language …” (179); “[A]s far as illocutionary forces are concerned there are five and only five fundamental types and thus five and only five illocutionary ways of using language” (52).
14“The Things We Do With Words: Ilongot Speech Acts and Speech Act Theory in Philosophy”, Language in Society 11, no. 2 (1982): 203–237.
15Martina T. Faller, Semantics and Pragmatics of Evidentials in Cuzco Quechua, Ph.D. dissertation submitted to the Department of Linguistics, Stanford University (2002); available at http://personalpages.manchester.ac.uk/staff/martina.t.faller/documents/Thesis.pdf.
16David W. Fleck, “Evidentiality and Double Tense in Matses”, Language 83, no. 3 (2007): 589–614.
17Ibid., 596. Obviously, this English translation carries implicatures that (one hopes) weren’t present in the original. Whether the Matses answer can be translated without such implicatures is a difficult question.
18Faller, Semantics and Pragmatics. A similar proposal has been made for the reportative evidential in Korean: Kyung-Sook Chung, “Korean Evidential and Assertion”, Proceedings of the 25th West Coast Conference on Formal Linguistics, ed. Donald Baumer, David Montero, and Michael Scanlon (Somerville, MA: Cascadilla Proceedings Project, 2006), 105–113.
19For an attempt to explain the use of evidentials (among other linguistic phenomena) in terms of cultural values, see Daniel L. Everett, Don’t Sleep, There Are Snakes: Life and Language in the Amazonian Jungle (New York: Pantheon, 2008). As Daniel B. Velleman has pointed out to me, some English utterances come close to having a presentational force: “Take, for instance, the claim that he’s out hunting.” Yet even this utterance differs from the Quechua, since its fundamental force is directive, so that it has a world-word direction of fit. The response “No, let’s not consider it” would be in order.
20Elinor Ochs Keenan, “The Universality of Conversational Postulates”, Language in Society 5, no. 1 (1976): 67–80. For an opposing argument, see von Fintel and Matthewson, “Universals in Semantics”, 189. Von Fintel and Matthewson say that Keenan’s argument is, to their knowledge, the only attempt to challenge the universality of Grice’s maxims. I would argue that the examples I am about to cite in the text — examples of conventional uncooperativeness in conversation — have similar implications for Grice.
21See also the quotation from Geertz, at note 28, below.
22Keenan, “The Universality of Conversational Postulates”, 73. As Daniel B. Velleman has pointed out to me, these utterances are not counterexamples to the Gricean maxim if they intentionally flout it in order to establish the speaker’s superiority — as if to say for example, “Someone is here, but I’m not going to tell you who it is.” But Keenan’s description indicates that uninformativeness is conventional in the language, not exceptional in a way that would convey such a message.
23See Richard F. Gombrich: “Lying is bound to be frequent in a culture much concerned with the preservation of status […] and dignity […] — saving face; the most trivial matter which might in any way appear discreditable to the speaker is concealed almost as a matter of course” (Precept and Practice: Traditional Buddhism in the Rural Highlands of Ceylon [Oxford: Clarendon Press, 1971], 263).
24Rachael M. Harris, “Truthfulness, Conversational Maxims and Interaction in an Egyptian Village”, Transactions of the Philological Society 94, no. 1 (1996): 41.
25The villagers say on the one hand, “We live with two faces”, and on the other, “The English don’t lie” (ibid., 35 and 43, respectively). Here is a similar report, of a remote Greek village where honor and prestige are similarly valued:
In the village the word for lies, psemata, is used much more freely, with less emotional intensity, and with a milder pejorative connotation than Americans use the English word. “Let’s tell a few more lies and then go home,” a man once remarked jovially near the end of a social evening. To accuse someone of mendacity is not the gross insult it is in the United States; it may be meant as a statement of fact in a situation in which, in village expectation, it would not be unusual for a person to attempt some deception.
Ernestine Friedl, Vasilika: A Village in Modern Greece (New York: Holt, Rinehart and Winston, 1962), 80. See also Juliet du Boulay, “Lies, Mockery and Family Integrity”, in Mediterranean Family Structures, ed. J.G. Peristiany (Cambridge: Cambridge University Press, 1976), 389–406.
26Gunter Senft, “The Case: The Trobriand Islanders vs H.P. Grice: Kilivila and the Gricean Maxims of Quality and Manner”, Anthropos 103 (2008): 139–147.
27Eve Danziger, “The Thought that Counts: Interactional Consequences of Variation in Cultural Theories of Meaning”, in Roots of Human Sociality: Culture, Cognition and Interaction, ed. N.J. Enfield and Stephen C. Levinson (New York: Berg, 2006), 260. References omitted.
28The Religion of Java (Glencoe, IL: The Free Press, 1960), 246 (interpolations are mine).
29Michael Gilsenan, “Lying, Honor, and Contradiction”, in Transaction and Meaning: Directions in the Anthropology of Exchange and Symbolic Behavior, ed. Bruce Kapferer (Philadelphia: Institute for the Study of Human Issues, 1976). On the meaning of ‘kizb’, see also Harris, “Truthfulness, Conversational Maxims and Interaction”.
30See Eve E. Sweetser, “The Definition of Lie: An Examination of the Folk Models Underlying a Semantic Prototype”, in Cultural Models in Language and Thought (Cambridge: Cambridge University Press, 1987), ed. Dorothy Holland and Naomi Quinn, 62. Sweetser claims, on the basis of fairly narrow cross-cultural evidence, that there is a core concept of lying that is universal.
31Ronald Hingley The Russian Mind (New York: Charles Scribner’s Sons, 1977), 78.
32Ibid., 87.
33This passage can be found in “Something About Lying”, from The Diary of a Writer, trans. Boris Brasol (New York: George Braziller, 1954), 133–142. I have substituted Hingley’s translation (83). Hingley translates Dostoyevsky’s title as “A Word or Two About Vranyo”.
34Here are a few further examples. The linguist Anna Wierzbicka writes, “There are many languages which have no exact equivalent of the word warning and which have, instead, words for modes of communication which have no equivalents in English. For example, Japanese has the word satosu, which combines some of the components of the English concept codified in the word warning with some other components: an assumption that the speaker has authority over the addressee, the intention of protecting the addressee from evil, and good feelings toward the addressee […]. In English, the assumption of authority is encoded in verbs such as order and forbid, but it is never combined (lexically) with the intention to protect. […] English doesn’t have any verb which would combine authority, responsibility, and care […].” (“A Semantic Metalanguage for a Crosscultural Comparison of Speech Acts and Speech Genres”, Language in Society 14, no. 4 (1985): 494). As Wierzbicka remarks, it may make no sense to talk of “‘questions in Eskimo’, ‘commands in Burundi’, or ‘blessings and curses in Yakut’”, because “English words such as question, command, or blessing identify concepts which are language-specific” (Wierzbicka, “All people eat and drink”, 492).
35Keith H. Basso, “‘To Give Up on Words’: Silence in Western Apache Culture”, Southwestern Journal of Anthropology 26, no. 5 (1970): 213–230. See also Karl Reisman, “Contrapuntal Conversations in an Antiguan Village”, in Explorations in the Ethnography of Speaking, ed. Richard Bauman and Joel Sherzer (Cambridge: Cambridge University Press, 1989), 112–113; and Ned Searles, “‘Why Do You Ask So Many Questions?’: Dialogical Anthropology and Learning How Not to Ask in Canadian Inuit Society”, Journal for the Anthropological Study of Human Movement 11, no. 1 (2000): 47–64.
36See Herbert H. Clark, “Social Actions, Social Commitments”, in Enfield and Levinson, Roots of Human Sociality, 126–150.
37“Walking Together: A Paradigmatic Social Phenomenon”, Midwest Studies in Philosophy 15 (1990): 1–14. I discuss this phenomenon in “How to Share an Intention”, Philosophy and Phenomenological Research 57, no. 1 (1997): 29–50, reprinted in The Possibility of Practical Reason (Ann Arbor, MI: MPublishing, 2009).
38Roger C. Schank and Robert P. Abelson, Scripts, Plans, Goals, and Understanding: An Inquiry Into Human Knowledge Structures (Hillsdale, NJ: Lawrence Erlbaum Associates, 1977). Schank and Abelson use the term ‘script’, which strikes my ear as implying that actions and utterances are mandated with more specificity than Schank and Abelson actually have in mind. I prefer the term ‘scenario’, which suggests a greater degree of indeterminacy, leaving room for improvisation. ‘Scenario’ is the term that was used for the standard plot outlines on which performers improvised in the Commedia dell’Arte tradition, and it was adopted by some of the originators of Chicago “improv” theater. (See R. Keith Sawyer, Improvised Dialogues: Emergence and Creativity in Conversation [Westport, CT: Ablex Publishing, 2003], 20 ff.) Where Schank and Abelson speak of scripts, and I speak of scenarios, Erving Goffman speaks of “routines” (The Presentation of Self in Everyday Life [New York: Anchor Books, 1959], 16 et passim). Sawyer discusses the variable specificity of scripts, scenarios, or routines in his Creating Conversations: Improvisation in Everyday Discourse (Cresskill, NJ: Hampton Press, 2001). For a recent philosophical discussion of scripts, see Cristina Bicchieri, The Grammar of Society: The Nature and Dynamics of Social Norms (New York: Cambridge University Press, 2006), 93 ff.
39“This is a stickup!” Did robbers actually say this back in the day? Was it said only in the movies? Or was it invented for use on the radio, where the audience couldn’t see what was happening?
40See, e.g., Amos Tversky and Daniel Kahneman, “The Framing of Decisions and the Psychology of Choice”, Science 211, no. 4481 (1981): 453–458.
41Weighing Goods (Oxford: Blackwell, 1991), Chapter 5. See also “Can a Humean Be Moderate?”, in Value, Welfare, and Morality, ed. R.G. Frey and Christopher W. Morris (Cambridge: Cambridge University Press, 1992), 51–73, p. 58. For a discussion of this phenomenon, see my paper “The Story of Rational Action”, Philosophical Topics 21, no. 1 (1993): 229–253; reprinted in The Possibility of Practical Reason.
42Note that ‘sightseeing’ is not what moral philosophers call a thick concept, since it is evaluatively neutral. To offer someone a day of sightseeing is neither to recommend nor to disparage the option. Yet the presence of ‘sightseeing’ in our practical repertoire has some practical import simply in virtue of defining the option in the first place — in virtue, that is, of constituting it as a doable. The notion of thick descriptions was introduced by Gilbert Ryle, “The Thinking of Thoughts: What is ‘Le Penseur’ Doing?”, reprinted in his Collected Essays 1929–1968: Collected Papers Volume 2 (London: Hutchinson, 1971), 480–496. See also Clifford Geertz, “Thick Description: Toward an Interpretive Theory of Culture”, Chapter 1 of The Interpretation of Cultures (New York: Basic Books, 1973), 1–31.